VPN Glossary: Subscriptions, Nodes, Route Types, Protocols & Routing Rules Explained

A clear introduction to the terms beginners encounter most: subscription links, nodes, IEPL dedicated lines, relayed and direct routes, common protocols, split tunneling, and global mode.

How Subscription Links Differ from Plan Subscriptions

A plan subscription describes how a service is billed; a subscription link tells your client where to fetch its configuration. When you import a link, the client typically retrieves node addresses, connection parameters, and authentication details. Some configurations also include groups or rules. Updating a subscription fetches the configuration again—it does not renew your plan. Conversely, after renewing, you may need to refresh the client if it is still using an old address.

A node is a connection profile you can select in your client. When you connect, the client first contacts the server specified in the profile. Websites generally see the outbound address that makes the final connection to them. The entry and exit points may be on the same server or on separate ones. So the number of nodes in a list does not necessarily equal the number of independent exit points.

Route Types: Direct, Relayed, and IEPL Dedicated Lines

When reading route labels, first distinguish whether they describe the access method or the conditions on a particular segment. Direct and relayed describe whether the provider has configured an intermediate relay. IEPL describes a segment that uses a dedicated Ethernet line. A relayed setup can also use a dedicated line between its entry and exit points, so these labels are not mutually exclusive tiers.

What route labels mean and how to assess them
Label Route described What to check in practice
Direct The client connects straight to the exit server, without an intermediate relay configured by the provider. Public internet traffic can still pass through multiple routers. Check latency fluctuations on your usual network and at typical times, and see whether the target app works reliably over time.
Relayed The client connects to an entry server first, which forwards traffic to an exit point elsewhere. A reachable entry server only confirms that the first segment works. You should also check how the connection performs from the exit to the target website.
IEPL dedicated line An international Ethernet private line, often used in route descriptions to identify the transmission segment between entry and exit points. Check which segment the dedicated line covers. The connection from your local network to the entry point, and from the exit point to a website, may still use the public internet.

Don’t rank routes by the latency number shown in your client alone. Some checks measure only the node’s entry point, while others request a test URL through the node, so the results aren’t directly comparable. To choose a route, use the same device and network at roughly the same time, open the same target app, and compare startup speed, sustained transfers, and interruptions.

Proxy Protocols and Client Compatibility

Protocols define how clients and servers establish connections, verify identity, and transfer data. Nodes in the same region may use different protocols. A successful import only means the client read some configuration; it doesn’t guarantee that your current version supports every protocol, transport method, or security parameter in it.

  • Shadowsocks: An encrypted proxy protocol. Common configurations use AEAD encryption; both ends must use matching encryption methods and keys. Some nodes also require a client that supports a specific plugin.
  • VMess and VLESS: These are different proxy protocols. VMess has its own authentication and data-encryption mechanisms; VLESS does not encrypt payloads on its own and is often paired with a security layer such as TLS or REALITY. Changing the protocol name in your client does not let you reuse the original connection parameters.
  • Trojan: Proxy traffic is typically carried inside a TLS connection. The server certificate, the name the client uses to verify it, and the authentication details must all match the configuration. Disabling certificate verification should not be treated as a routine fix.
  • Hysteria2 and TUIC: Both are based on QUIC and typically rely on UDP. If your current network cannot forward UDP properly, a node may import successfully but still fail during the handshake. These are different protocols, and their parameters are not interchangeable.

Also check how the client handles app traffic. On Windows, the system proxy usually affects only software that follows the system proxy settings. Clients with TUN support can cover more apps, though routing and exclusion rules still determine the actual scope. Android clients often use the system VPN interface to handle traffic. Options with the same name may behave differently across platforms; check the Guides to confirm how traffic is handled on your device.

Split Tunneling, Global Mode, and DNS

Split-tunneling rules determine which connections use the proxy, connect directly, or get blocked. Rules can match domains or IP ranges; some clients can also match apps. Many clients evaluate rules from top to bottom and stop at the first match. If a broad rule comes first, exceptions listed later may never take effect. If an internal service should connect directly, check whether the internal network rule actually matches rather than just checking whether the “Direct” group is selected.

Global mode usually sends only connections handled by the client through the selected exit. It does not necessarily route every app or DNS request on the device through the tunnel. If a website works in global mode but not in rule mode, check which rule matches and how the domain is resolved. If it fails in both modes, check the node handshake, the target service, and whether the app’s traffic is being handled.

DNS translates domain names into addresses. A DNS leak occurs when a request that should be resolved through the proxy instead goes to a local resolver. A changed exit address visible to a website does not mean DNS took the same route. When troubleshooting, check the client’s DNS settings, encrypted DNS in your browser, and system DNS settings. If the client provides connection logs, check both the DNS resolution path and the rules that matched the connection. Local domains configured for direct resolution do not need to be treated as leaks.

How to Fix Connection Issues After Importing

  1. First, check whether the subscription updated successfully. If the list is empty, check whether your client supports the subscription format provided by the dashboard and whether any errors appeared while fetching the configuration. Changing the node protocol won’t fix an import problem.
  2. Next, check whether the node can complete a handshake. If the list updated but the connection fails, verify that your client supports the node’s protocol and transport method. If the same node connects on a different network, compare the network conditions first; avoid changing multiple connection parameters at once.
  3. Finally, identify which app is affected. If your browser works but a desktop app doesn’t, check whether the client is handling that app’s traffic. If only some domains fail, check the matching rules and DNS. If the connection starts but becomes unstable during use, compare other routes in the same app.

Troubleshoot in layers: The subscription determines where the configuration comes from; the protocol determines whether the two ends can connect; the route affects performance once connected; and routing rules determine which traffic uses that route. Finding the layer where the problem occurs is more effective than changing the node, protocol, and mode all at once.

Start Free